Nostr was always a weird beast, it is neither a shameless message passing protocol like Activity Pub, nor is it highly consistent through centralisation like At protocol, nor is it highly consistent through proof of publication (DA) like Farcaster.
Instead, it is more like the Bitcoin mempool than anything else, a broadcast network that is too unreliable for good UX, but its main goal is to just let things eventually slip through a supposed censorship. Admirable really, but it doesn't work any better than sharing links in group chats and emails until they reach their target audience.
I am afraid that the only viable way for censorship resistant yet reliable broadcasting is a scalable proof of publication ephemeral decentralized ledger...
Luckily that is also the endgame solution for scaling payments so it will eventually happen, but I am not sure anyone cares to broadcast their tweets so much as to pay for it, especially that it only guarantees that the data is available, but it doesn't guarantee that anyone will read it or see it at all, anymore than Bitcoin guarantees that a transaction is going to appear on people's wallets if they don't ask for it.
Which brings me to the question... Is this just a note taking app?
Replies (15)
Yes.. the answer is yes of course
Most people here do not their stuff to be perpetually saved on a ledger.
its a note taking app absolutely. But each note is verifiable and sent accross the www.
then the content is addressable and you can build all sort of weird stuffs with it.
was greatly missed tbh
This is a common misconception that a proof of publication ledger means data that magically can't be forgotten... That only applies to Bitcoin, for very specific reason.
The reality is;
1. A proof of publication is not about perpetual availability it is just about consensus on what is everyone is seeing in a given window of time.
2. Nostr doesn't do anything to stop perpetual archival beside just being not important to archive.
If you really want you can force the archiving thing on Nostr without its consent by simply crawling the top most relays and timestamping that with OTS.
But that wouldn't be a proof of publication, because it is not signed by the relays, so no evidence that the data existed on popular relays that people query.
So now you have all the downsides of archiving (user signed events) but no proof of publication (proof that most people should have seen this in time).
In fact for a proof of publication to scale it has to prune history or in other words it has to only prove publication temporarily not perpetually... But again, archiving is another layer that is equallu possible and simple with or without proof of publication.
You missed the point. The goal is not even deal with any of that stuff.
I don’t want to deal with any of that stuff.
No you missed my point, the stuff that I said Nostr isn't dealing with is relative consistency that allows apps to give users an experience they expect from social media apps.
So saying that you don't want to deal with the price of perpetual ledgers is irrelevant... Bluesky doesn't have any kind of ledgers at all, they solve it with extreme centralisation, similar to Primal, which is why I literally have to come to Primal every time I need to actually see if I have notification faster than waiting for 2 work days and then miss many.
I am not saying this is something everyone should want, I repeatedly said that this is a problem that I don't care about much and prefer direct chat apps.
But if you want to build a broadcast network, with the UX of social media, then sorry, but having a solution for reliable and quick propagation of data to the indexers is not optional.
When you say that most Nostr folks don't care about that, then yes I believe you, because like me they just use it as a note taking app with little expectations of reach or engagement.
But I figured it is useful to explain these choices and tradeoffs so I wrote about them.
Direct chat = a same relay use case.
For the reasons you mention.
I literally can't see your likes or replies until I open Primal, and even then the replies are more reliable if I open the thread ... Which is insane because this could have been a direct message like email or ActivityPub it has no business being Pull only flow... But here we are.
But yes if we must pull only, well at least pull from fewer relays, which is easier with either centralisation or sharding by context, but if you want to see threads notifications from your home page let alone get push notifications, there is simply no alternative to fewer sources of data.
People really don't understand this, more relays is bad
They don't understand this because relays are acting as data store, and obviously more data stores is good.
Yup. Also, it makes me lol that primal is your most reliable notification solution.
(#nostrudel wins for me, on mobile too)
By the way this is why DHTs work; they structure relays/nodes so each bunch (~8-20) are the rendezvous of specific topic.
I think the Outbox Inbox framing is good solution, so I only read notification from few relays...
However this doesn't work for global open chats because... Where the fuck should I send my replies to a thread with 300 participants? What about lurkers who still engaged? Do I send the reply to everyone relays?
Well in that case you better hope for relays centralisation
But you see, doing full #outbox is the new flex around here.
So now, 21 million relays is the bare minimum if you want to keep your grant, sir.
preach
Yes, you should send to all the inbox relays of the 300 participants. And that works on Amethyst just fine.
